Medical errors are common, harm patients and increase the cost of healthcare in the US by billions of dollars each year. Many of these errors are preventable. However, in order for healthcare providers to learn from their errors and prevent them in the future, they must first be aware that they have occurred. This feedback process is vital, but does not occur reliably. The implications of this deficiency are especially important for physician trainees, when diagnostic approaches and treatment decisions are learned. As a consequence of regulations to reduce resident work hours, physician training has increasingly become shift work and the care of patients is often handed off from one physician to another. These handoffs currently prevent many trainees from receiving feedback regarding the effects of their clinical decisions because these effects often take time to become evident. The objective of this project is to address this lack of feedback by developing an automated patient outcome feedback system within an electronic health record (EHR). We hypothesize that this system will enable physicians to learn the outcomes of their patients after handoffs, allowing them to improve their decision-making and reduce errors over time. The proposed project will include three main aims to address the lack of patient outcome feedback after handoffs. In Aim 1, we will establish the extent of the problem by measuring how often emergency medicine and internal medicine physician trainees currently view their patients' charts after handoffs. In Aim 2, we will create a set of electronic triggers that will identify cses in which physicians may have made errors. Finally, in Aim 3, we will develop and pilot test a system within an EHR that will automatically provide physicians with a report of the outcomes of their patients after handoffs. At the conclusion of our proposed studies, we will have created a methodology to improve the frequency and reliability of patient outcome feedback to emergency medicine and internal medicine physician trainees following handoffs. If successful, the system will result in more accurate initial diagnoses and more effective treatment decisions, leading to fewer medical errors.